TexerII & Colormap

This is a request for unconed, two things I would like to see in your APE's:

TexerII: The option to have a black and white mask for the image used, so that one can do replace blended effects

Color Map: The ability to pick the color map number with registers, and maybe the map cycling too. This would be useful for mouse control presets where I would like to have the option of picking the color map in a user menu (I know it can be hacked around with coeable effect lists but its slower and uglier)

They arent particularily complex requests, but things that I think would make the APEs more useful.

i really love the trans texer for that ability texer2 doesn't have 🙂 thats why sometimes i still use it if n is not above 1024. with it, you can also render adjustable blend and that results in really nice holographic effects sometimes 🙂

and i just agree with the second. though i think i will never really use it, too much work now.

How would the alpha mask affect bilinear filtering?

how about also accepting non-integers for the color maps?
(for a start)
if you want to invest further time in that, you could even allow the user to set an alpha value for each color map so that they can be blended together in whatever way you like

Originally posted by UnConeD
How would the alpha mask affect bilinear filtering?

?

It wouldn't, I assume you mean bilinearily resizing the images for sizex and sizey. You'd get some blurring but it wouldnt really notice, I guess you could bucibicly resize them if you thought it would, but computationally its not really worth it.

Tell me if I've completely got the wrong end of the stick here.

You should add a feature like: "Treat X,X,X color as transparent" so that people like hboy who still use old technology could upgrade 😛

That's the difference between texer & texer II, other than texer II being vastly superior, but it just doesn't have that one small thing.

...and colour filtering
...and resizing
...and having more than 1024 points
...and being directly codable

I'll assume you grouped that all in being "vastly superior"

You catch up pretty quickly PAK 🙄

Yes, it might not be too "clear" what vastly superior ment, but thanks to you and your incredible skills to point-out-things it's not unclear anymore!



wanker...

bleh, i like and use both 🙂

I think maybe UCDs question is a lazy rhetorical attempt to get out of APE coding. 😉

I can't see whats so hard about implementing both alpha and bilinear together. bilinear filter the rgb and bilinear filter the alpha then final=background*(1-alpha)+rgb*alpha

Other problems? You get a blurry line instead of a clean one which could spoil an effect. Solution. Turn off bilinear.

An alpha image would be useful.

I dont think filtering the rgb then the alpha is really necessary, you can just multiply them and filter the result.

you can't do that

think about it. if you have a circle for instance that you want to replace blend only the circle and bilinear filter the whole jobby. you need to bilinear the alpha seperately for the final combine so that it blends the edges... otherwise you get ugly almost black edges on the circle where you blend rgb*alpha and use the original channel.

hmm, I was thinking you could multiply the scene by the mask, add the rgb then do the filtering afterwards but that doesnt really make any sense now I think about it, cos the filtering is part of the initial resizing before the mask is even applied. Never mind.

btw, you could add nifty +/- buttons for your code boxes 🙂

Ability to rotate at Texer II would be nice 🙂

write your own rotation code

(if you mean rotating sprites, that is not possible and would kill the framerate if implemented... use a DM or movement instead)

UnConeD said rotating particles was on his "to do" list, so i guess it wouldn't be such a fps hog as you say TomyLobo.

vertex shading was on my todo list for triangle.ape too, but i scrapped it because it wasnt really possible without a big fps hit

Optional features are a zero frame rate hit, it never hurts to have the option if people want it.

want the source? :P
be warned though, it's 400 lines of asm :P

I eat 400 lines of assembly for breakfast, with a cup of tea (milk no sugars please).

I'm too damn busy to work on it unfortunately, but if you send it to me I might get around to it eventually.

Animated GIF support would be COOL 😁

I can hack in animation support by simply chopping up the image into vertical tiles. There will never be such a thing as animated GIF support.

In any case, I really don't have the time to implement this. Pak: the "optional" feature of alpha blending (or rotation for that matter) would require a new code path of asm code.

hey unconed. why isn't it possible to make avs "open source" or something? It's more than probably that no new avs version comes out ever, but there are still many people interested in AVSing. It would only be fair to let them develop it further by themselves! (sorry for bad english, but im still kind of drunk 🧟 )

Wow, someone else beside me, posts drunken posts 😁

just *had* to say it

yeah! It's kind of funny, isn't it? 🙂

if i remember correctly, someone already asked nullsoft to release the source, but they refused.
never mind though, we have something in the fridge.. err sleeve 🙂
first results will probably not be available before next year, but that's still earlier than an avs update (which is expected at about the time when hell freezes over)

but why did they refuse to release the source-code? I mean sometime avs is so much out-of-date, that they probably won't add it to winamp anymore. At least it would be an advantage, if they could bring out that new vis together with winamp then, isn't it?

And what about asking Justin instead of Nullsoft?

Is he the legal owner of avs? Or does nullsoft own the rights? Not that justin gets sued when he gives out the code...